Comparing pedagogical approaches for the acquisition and long-term robustness of the control of variables strategy

Authors:
Michael A. Sao Pedro;Janice D. Gobert;Juelaila J. Raziuddin
Affiliations:
Worcester Polytechnic Institute, Worcester, MA;Worcester Polytechnic Institute, Worcester, MA;Worcester Polytechnic Institute, Worcester, MA
Venue:
ICLS '10 Proceedings of the 9th International Conference of the Learning Sciences - Volume 1
Year:
2010

Citing 2
Cited 2

The Andes Physics Tutoring System: Lessons Learned

International Journal of Artificial Intelligence in Education
Accelerated Future Learning via Explicit Instruction of a Problem Solving Strategy

Proceedings of the 2007 conference on Artificial Intelligence in Education: Building Technology Rich Learning Contexts That Work

The dynamics between student affect and behavior occurring outside of educational software

ACII'11 Proceedings of the 4th international conference on Affective computing and intelligent interaction - Volume Part I
Leveraging machine-learned detectors of systematic inquiry behavior to estimate and predict transfer of inquiry skill

User Modeling and User-Adapted Interaction

Quantified Score

Hi-index	0.00

Visualization

Abstract

This study compared three pedagogical approaches on the acquisition and robust understanding of the control of variables strategy (CVS). In Sao Pedro et al. (2009), we showed that two direct learning conditions (with and without reification) significantly outperformed the discovery learning condition for constructing unconfounded experiments starting from an initially multiply confounded experimental setup. In the study described here, we retested 57 students six months later on constructing unconfounded experiments in a virtual ramp environment, solving problems requiring CVS, and explaining CVS. Collapsing over time, we found that the direct+reify condition had more robust learning than either the direct-no reify or discovery learning conditions on constructing unconfounded experiments. At the delayed posttest, we found a strong trend favoring the direct+reify condition over the other conditions as measured by tasks designing unconfounded experiments starting from a multiply confounded state.